New app checks fruit and vegetables for chemical residues

With a new app created by Fraunhofer research you can display the ingredients of an object. The possibilities of the application are numerous: for example, you can check for residues on an apple.

The ability to check an apple for chemical residues with a smart phone will soon be possible. An apple can be labeled organic, but the customer doesn’t know if the apple is still exposed to chemicals. And according to the salesperson - 'the car never had any damage'. Often you have to trust these statements, without the opportunity to check them. With the ‘HawkSpex mobile’ application created by Fraunhofer Institute for Factory Operation and Automation IFF in Magdeburg, consumers will have the opportunity to check for themselves. The approach is simple: you take your smart phone, open the app, aim the camera at the object, for instance an apple, and get the relevant information, for instance if there is any chemical residue on the apple.

There are already apps available that can provide this information, however the user needs to equip the phone with extra parts, for example a prism in front of the camera, which is expensive and impractical. And the extra parts limit the use and design of the smart phone. “The special feature of our app is that the user only needs the camera, which is already integrated into his smart phone,” states professor Udo Seiffert, Expert Group Manager at IFF.

How did the engineers, working with project manager Dr. Andreas Herzog, manage this feat without a prism? Scanning like this usually requires a special hyperspectral camera. It can adjust each time to the different colored light and measures how much light and which color is reflected by an object. This way the camera generates a complete spectral image of the object. The spectral image is combined with a mathematical model to extract a wide range of information about an object, for example its constituents.

“Hyperspectral cameras aren’t integrated in smart phones, so we simply reversed this principle,” explains Seiffert. “The camera is a three-way sensor, one that scans every wavelength and illuminates an object with different colored light.” And instead of the camera measuring the intensity of different colors, the app uses the phone display to successively illuminates the object with series of different colored light each for a fraction of a second. When the display shines only red light on the object, the object can only reflect red light – and the camera can only measure red light. Smart algorithms enable the app to compensate for the limited computing power of the phone, as well as the limited performance of the camera and display.

The first beta version of the app is ready and patent application is pending. Though, before it can be released for private use engineers are still developing a variety of applications. The app learns, with the use of reference scans, how to analyze whether apples contain pesticides. Seiffert hopes that the “HawkSpex® mobile” app will be launched at the end of 2017.

Reference scans are not always needed, because some problems only require measuring different distributions rather than specifying individual constituents. For instance, in the case of the car example, the app compares the paint to determine whether it is exactly the same color everywhere or if it has been touched up anywhere.

“There are so many possible uses that the market will be very enthusiastic,” says Seiffert with certainty. That is why the development engineers are relying on an approach modelled on the online encyclopedia Wikipedia. “Once the app is available, active users will be able to contribute and create new applications, for instance, a test function that checks the pesticide exposure of lettuce. They teach the system how to deal with such problems,” says Seiffert. They will use the app to scan different types of treated and untreated lettuce and send the data to the Fraunhofer IFF. Engineers verify the measurements and release the function to all users.

The app has extremely interesting commercial potential and can be used in sectors that wouldn’t benefit from high precision scanners. The list of possibilities is nearly endless; it can be use for food, cosmetic products or even in agriculture. Farmers, for instance, could easily find out if their crops have sufficient nutrients or if fertilizer is needed.

For more information:

Fraunhofer Institute for Factory Operation and Automation IFF